An integrated framework for trademark image retrieval using gestalt features and CMM neural network

This paper describes a novel framework for trademark image retrieval using gestalt features and correlation matrix memory (CMM) neural networks. Its feature extraction process is aimed at capturing features from different analytical levels of images using gestalt feature extraction methods. The proposed search engine of the system consists of associative memory nodes connected by information channels which convey symbolic messages. Symbolic information stored inside the system is obtained from the earlier mentioned feature extraction process. In this paper, we describe the structure of the system, training process and the process adopted to measure the perceptual similarity of trademark images. Finally, we present results obtained during evaluation of this framework.